diff --git a/lib/regex_helper.rb b/lib/regex_helper.rb
index 5edfa99f4..46830d33f 100644
--- a/lib/regex_helper.rb
+++ b/lib/regex_helper.rb
@@ -4,7 +4,9 @@ module RegexHelper
   # the following regext needs atleast one character which should be
   # valid unicode letter, unicode number, underscore, hyphen
   # shouldn't start with a underscore or hyphen
-  UNICODE_CHARACTER_NUMBER_HYPHEN_UNDERSCORE = Regexp.new('\A[\p{L}\p{N}]+[\p{L}\p{N}_-]+\Z')
+  # \z (not \Z) anchors strictly to the end of the string -- \Z would also accept a single
+  # trailing "\n", which would let a title like "hello_world\n" pass this validation.
+  UNICODE_CHARACTER_NUMBER_HYPHEN_UNDERSCORE = Regexp.new('\A[\p{L}\p{N}]+[\p{L}\p{N}_-]+\z')
   # Regex to match mention markdown links and extract display names
   # Matches: [@display name](mention://user|team/id/url_encoded_name)
   # Captures: 1) @display name (including emojis), 2) url_encoded_name
diff --git a/spec/models/label_spec.rb b/spec/models/label_spec.rb
index ee5e7b108..ab5a00d1a 100644
--- a/spec/models/label_spec.rb
+++ b/spec/models/label_spec.rb
@@ -26,6 +26,13 @@ RSpec.describe Label do
       expect(label.valid?).to be true
     end
 
+    it 'would not let you use a title with a trailing newline' do
+      # Regression test: the format validator used to anchor on \Z instead of \z, and \Z tolerates
+      # a single trailing newline, so 'hello_world' + "\n" incorrectly passed validation.
+      label = FactoryBot.build(:label, title: "hello_world\n")
+      expect(label.valid?).to be false
+    end
+
     it 'converts uppercase letters to lowercase' do
       label = FactoryBot.build(:label, title: 'Hello_World')
       expect(label.valid?).to be true